Advanced AI Technologies
Oracle broadens cloud AI hardware portfolio
Oracle expanded its Nvidia partnership and plans to offer Google, Cerebras and SambaNova AI processors through its cloud platform.
Oracle announced an expansion of its agreement with Nvidia to deploy Nvidia’s Vera Rubin systems at multi-gigawatt scale. Oracle is also broadening the processors available through Oracle Cloud Infrastructure, including Google tensor-processing units and inference chips from Cerebras Systems and SambaNova. The move gives cloud customers access to a wider selection of specialized hardware for training and running AI models. Different processors can be suited to different workloads, including high-performance training, inference and cost-sensitive deployment. Oracle’s approach reflects increasing competition among cloud providers to secure AI computing capacity and reduce dependence on a single chip supplier. Multi-gigawatt deployment plans also illustrate the scale of infrastructure now being planned for frontier AI. For enterprises, a broader hardware portfolio could improve flexibility when selecting platforms for demanding machine-learning applications. The announcement is part of a wider transition in cloud computing, in which accelerators and specialized AI systems are becoming central components of digital infrastructure.
Original source: InformationWeek